Scholarlink

<p>Estimating heterogeneous causal effects is a common problem in statistics and econometrics studies. Model averaging has been a powerful tool for estimating conditional average treatment effect (CATE). Within this context, we propose a model averaging method by jackknife criterion of varying-coefficient models (VCMs) for estimating heterogeneous causal effects. We provide some theoretical justification for our model averaging approach as well as establish the asymptotic optimality property, weight convergence property and asymptotic normality. In the simulation parts, we examine the finite-sample performance of our estimator and compare it with several other model selection and averaging methods. We illustrate our method using a real-data dataset from a labor skills training program.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Model averaging by Jackknife criterion of varying-coefficient models for estimating heterogeneous causal effects

  • 期刊论文

  • 发表日期：

    2025年7月14日

查看全文
  • Xiaowei Zhang,
  • Ziyu Wang,
  • Guangren Yang,
  • Xiyue Zhang

摘要

Estimating heterogeneous causal effects is a common problem in statistics and econometrics studies. Model averaging has been a powerful tool for estimating conditional average treatment effect (CATE). Within this context, we propose a model averaging method by jackknife criterion of varying-coefficient models (VCMs) for estimating heterogeneous causal effects. We provide some theoretical justification for our model averaging approach as well as establish the asymptotic optimality property, weight convergence property and asymptotic normality. In the simulation parts, we examine the finite-sample performance of our estimator and compare it with several other model selection and averaging methods. We illustrate our method using a real-data dataset from a labor skills training program.